Abstract
This paper investigate the fractional-order dynamics during the evolution of a Genetic Algorithm (GA). In order to study the phenomena involved in the GA population evolution, themutation is exposed to excitation perturbations during some generations and the corresponding fitness variations are evaluated. Three similar functions are tested to measure its influence in GA dynamics. The input and output signals are studied revealing a fractional-order dynamic evolution.

Abstract
An adaptive greenhouse climate controller was implemented to regulate the air temperature, humidity and carbon dioxide concentration, with the aim of achieving set-point accuracy and reduce energy consumption. An optimization algorithm, based on the minimization of a cost function, was used to tune a Proportional-Integral-Derivative controller. The cost function is computed over a future time horizon of one hour as a function of the errors between the predicted and desired outputs and the predicted energy demand. Since the controller must be able to predict the greenhouse climate, it was needed to employ recursive identification algorithms to estimate in real-time the parameters of the climate model. When compared with commercially available controllers, this adaptive controller proved to have better performance regarding set-point accuracy and energy consumption.

Abstract
The technique of genetic algorithms (GAs) is proposed as a means of auto-tuning PI Smith predictor controllers. The technique involves firstly using on-line data and the GA to identify a model of the process. Then the identified model, the GA and simulation methods, are used to off-line tune the PI Smith predictor controller, so as to minimise a time-domain based cost function. Finally, the genetically tuned controller is implemented on-line on the real process. The results of the auto-tuning of the PI Smith predictor controller technique are illustrated on a laboratory heat exchanger, and a comparison between this technique, and both the genetic PI auto-tuning technique and the Astrom PI auto-tuning technique are made.

Abstract
Artificial co-evolutionary techniques are proposed in a new and novel paradigm to solve the problem of designing a robust fixed PID controller for a plant with prescribed plant uncertainties. The co-evolutionary scheme used, involves generating two separate populations, one representing the controller and the other the plant. Two separate cost functions are then used in the co-evolutionary scheme to reflect the different goals of the two populations. The two populations are then co-evolved such that the population of plants, with the prescribed uncertainties, contains the set of difficult plants to control and a population of controllers emerges, which can control all these difficult plants effectively. The resulting paradigm not only results in a robust controller design but also produces a set of worst case plants. This co-evolutionary approach is illustrated through co-evolving a PID controller for a linear plant which has a set of prescribed uncertainties.

Abstract
The aims of the present study were: to identify the factors which are able to explain the performance in the 200 meters individual medley and 400 meters front crawl events in young swimmers, to model the performance in those events using non-linear mathematic methods through artificial neural networks (multi-layer perceptrons) and to assess the neural network models precision to predict the performance. A sample of 138 young swimmers (65 males and 73 females) of national level was submitted to a test battery comprising four different domains: kinanthropometric evaluation, dry land functional evaluation (strength and flexibility), swimming functional evaluation (hydrodynamics, hydrostatic and bioenergetics characteristics) and swimming technique evaluation. To establish a profile of the young swimmer non-linear combinations between preponderant variables for each gender and swim performance in the 200 meters medley and 400 meters font crawl events were developed. For this purpose a feed forward neural network was used (Multilayer Perceptron) with three neurons in a single hidden layer. The prognosis precision of the model (error lower than 0.8% between true and estimated performances) is supported by recent evidence. Therefore, we consider that the neural network tool can be a good approach in the resolution of complex problems such as performance modeling and the talent identification in swimming and, possibly, in a wide variety of sports.

Abstract
Parallel manipulators have attracted the attention of researchers from different areas such as: high-precision robotics, machine-tools, simulators and haptic devices. The choice of a particular structural configuration and its dimensioning is a central issue to the performance of these manipulators. A solution to the dimensioning problem, normally involves the definition of performance criteria as part of an optimization process. In this paper the kinematic design of a 6-dof parallel robotic manipulator is analyzed. Three performance criteria are formulated and optimal solutions are found through a particle swarm formulation.